INDEX

Explanations

key architectural terms and metrics related to model performance in machine learning

oai_token-act-pair · gpt-4o-mini

New Auto-Interp

Configuration

fnlp/Llama-Scope-R1-Distill/800M-Slimpajama-0-OpenR1-Math-220k/L15R

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Hzfinfdu/SlimPajama-3B

Features

32,768

Data Type

float32

Hook Name

blocks.15.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

able

-0.14

ity

-0.14

ned

-0.14

-0.13

ono

-0.13

anean

-0.13

ables

-0.13

apel

-0.13

apest

-0.13

ape

-0.13

POSITIVE LOGITS

amma

0.16

DataExchange

0.15

 Mess

0.14

 Thornton

0.14

arro

0.14

ureka

0.14

åĽ´

0.14

enco

0.14

eller

0.13

deen

0.13

Activations Density 0.950%

key architectural terms and metrics related to model performance in machine learning

No Comments

No Known Activations